Seeing language from the DNA

[Jaska]

Ph2ter:

I don't belong to any of the groups made in your head.

Any? Not even in the scientific group? What is your methodical view, then?

I don't know what 'scientific' means to you.

The same it means to everyone else: that we stay in the reliable scientific methods and acknowledge that different disciplines can only study their own object. That we cannot use genetic methods to study the language any more than we can use methods of folk musicology to study the atmosphere of Mars. Do you agree with this?

[Gortaleen]

Ph2ter:

I don't belong to any of the groups made in your head.

Any? Not even in the scientific group? What is your methodical view, then?

I don't know what 'scientific' means to you.

The same it means to everyone else: that we stay in the reliable scientific methods and acknowledge that different disciplines can only study their own object. That we cannot use genetic methods to study the language any more than we can use methods of folk musicology to study the atmosphere of Mars. Do you agree with this?

Science is a process:
1. observation 
2. hypothesis about that observation 
3. test the hypothesis 
4. publish

Regarding DNA vs language we have an example observation.  The Kurgan hypothesis for the spread of Indo-European languages fits very well with DNA findings regarding the migration of Indo-European peoples. Any possible hypothesis for that observation?

I’m sure archaeologists can chime in too.

[Cejo]

I don't know what 'scientific' means to you.

The same it means to everyone else: that we stay in the reliable scientific methods and acknowledge that different disciplines can only study their own object. That we cannot use genetic methods to study the language any more than we can use methods of folk musicology to study the atmosphere of Mars. Do you agree with this?

Emphasis is mine.

To me, subjecting hypotheses to evidence from a broad range of disciplines is often critical in achieving the rigor demanded by the scientific method. Restricting inquiry based on discipline/method is in some regards antithetical to what scientific means to me. In my view, an interdisciplinary viewpoint helps when it comes to the inductive reasoning (hypothesis-forming) phase of the scientific method, and when trying to avoid cognitive biases when analyzing results. 

Basically, I evidently don't attach the same meaning to "scientific" as you do, though I am a physical scientist, which may skew my viewpoint.

[ph2ter]

Ph2ter:

I don't belong to any of the groups made in your head.

Any? Not even in the scientific group? What is your methodical view, then?

I don't know what 'scientific' means to you.

The same it means to everyone else: that we stay in the reliable scientific methods and acknowledge that different disciplines can only study their own object. That we cannot use genetic methods to study the language any more than we can use methods of folk musicology to study the atmosphere of Mars. Do you agree with this?

As Cejo already mentioned, I think that interdisciplinary studies are very important for many problems. And they are perfectly scientific.

[jdean]

As Cejo already mentioned, I think that interdisciplinary studies are very important for many problems. And they are perfectly scientific.

But surely interdisciplinary means people from different disciplines coming together, not excluding one because the rest think they can do a better job ?

[Anglesqueville]

If, then, the results of genetics, archaeology, and historical linguistics do not align, how is it possible for those disciplines to conduct a meaningful conversation about human history? First of all, they need to acknowledge their differences and avoid metaphors borrowed from one another. Celtic (*) is a well-defined linguistic term, and its application to archaeological cultures and to genes is a metaphor that is bound to cause confusion and misunderstandings.
A second means of meaningful exchange is to concentrate on windows of opportunity: if, say, a fairly uniform archaeological culture is seen to spread across a wide area, this is a window of opportunity for a language to have spread along with the culture (but by no means a necessity). If this happens to coincide with independently recovered data from historical linguistics, we may surmise that language and culture in this particular case spread in tandem. A third precondition to cross-disciplinary cooperation is that we abstain from propping up a weak linguistic argument with archaeological or genetic data and vice versa. An argument from linguistics should be able to stand on its own two feet and be viable by itself.

(*) Replace "Celtic" with what you want.

Peter Schrijver "Language contact and the origins of the Germanic languages", Routledge 2014

[ph2ter]

As Cejo already mentioned, I think that interdisciplinary studies are very important for many problems. And they are perfectly scientific.

But surely interdisciplinary means people from different disciplines coming together, not excluding one because the rest think they can do a better job ?

Sometimes some discipline can have a lower level of development. This doesn't mean that other disciplines need to forcibly comply with its results.
When all disciplines come to an agreement then it is real progress with fairly proven fact, but if they do not agree then this means that some of them are wrong.

[Jaska]

Cejo:

To me, subjecting hypotheses to evidence from a broad range of disciplines is often critical in achieving the rigor demanded by the scientific method. Restricting inquiry based on discipline/method is in some regards antithetical to what scientific means to me. In my view, an interdisciplinary viewpoint helps when it comes to the inductive reasoning (hypothesis-forming) phase of the scientific method, and when trying to avoid cognitive biases when analyzing results.

Interdisciplinary is of course important, too. But it requires by definition acknowledging the limitations of disciplines – it does not mean that language could be seen from the DNA or vice versa. Unfortunately many people who claim to be interdisciplinary are actually only unscientific, when they claim to achieve scientific knowledge by unscientific methods. The risk of cognitive bias is also present in interdisciplinary work.

Cejo:

Basically, I evidently don't attach the same meaning to "scientific" as you do, though I am a physical scientist, which may skew my viewpoint.

To me it seems that your definition agrees with mine. What do you think now?

Gortaleen:

Regarding DNA vs language we have an example observation.  The Kurgan hypothesis for the spread of Indo-European languages fits very well with DNA findings regarding the migration of Indo-European peoples. Any possible hypothesis for that observation? I’m sure archaeologists can chime in too.

The Kurgan model works for Late or nuclear Indo-European, but the question of Indo-Anatolian homeland is a bit more complicated – at least according to some geneticists. Archaeologists and linguists have had no problem deriving also Anatolian languages from the steppe, but some geneticists now propose that the Indo-Anatolian homeland was in the Caucasus Region.

However, these geneticists also consider it possible that the Indo-Iranian languages would have spread from there directly to the east, which is of course impossible from the linguistic point of view. This illustrates how difficult it is to see the language from the DNA: one cannot just observe the largest ancestry component shared by populations of the same language family, and claim that it is associated with the expansion of that language family. The result from such a method can easily be false. Only linguistics has reliable methods to study language.

[Jaska]

Ph2ter:

Sometimes some discipline can have a lower level of development. This doesn't mean that other disciplines need to forcibly comply with its results.

If by “development” you mean that some discipline does not offer unambiguous results to some question, then I agree.

Ph2ter:

When all disciplines come to an agreement then it is real progress with fairly proven fact, but if they do not agree then this means that some of them are wrong.

How could different disciplines disagree, even in theory? After all, they do not and even cannot study the same object. The results cannot disagree - only interpretations can.

Take the example from my previous comment:
Geneticist observe a large shared ancestry component, which they believe is related to the spread of a language family. Linguists say that such an interpretation is impossible, because linguistic evidence excludes such an interpretation. Still, there is no actual disagreement between the results of different disciplines: the large ancestry component is still there, and the linguistic results are still there. These results just cannot overrule each other - they both must be accepted. 

- Interdisciplinary conclusion is that this ancestry component cannot be associated with the spread of that language family.
- Unscientific conclusion is that one can ignore the linguistic results and claim to see the language better from the DNA.

[ph2ter]

Ph2ter:

Sometimes some discipline can have a lower level of development. This doesn't mean that other disciplines need to forcibly comply with its results.

If by “development” you mean that some discipline does not offer unambiguous results to some question, then I agree.

Ph2ter:

When all disciplines come to an agreement then it is real progress with fairly proven fact, but if they do not agree then this means that some of them are wrong.

How could different disciplines disagree, even in theory? After all, they do not and even cannot study the same object. The results cannot disagree - only interpretations can.

Take the example from my previous comment:
Geneticist observe a large shared ancestry component, which they believe is related to the spread of a language family. Linguists say that such an interpretation is impossible, because linguistic evidence excludes such an interpretation. Still, there is no actual disagreement between the results of different disciplines: the large ancestry component is still there, and the linguistic results are still there. These results just cannot overrule each other - they both must be accepted. 

- Interdisciplinary conclusion is that this ancestry component cannot be associated with the spread of that language family.
- Unscientific conclusion is that one can ignore the linguistic results and claim to see the language better from the DNA.

Some discipline can give wrong results, because it can be based on false premises and subsequent false conclusions. It can have false foundations. So, in such case the results/interpretations can disagree. And they can study the same object from different perspectives.

[jdean]

Getting back to language change without a related change in DNA, English is officially an administrative language of India

OFFICIAL LANGUAGE RELATED PART-17 OF THE CONSTITUTION OF INDIA

Notwithstanding anything in part XVII, but subject to the provisions of article 348, business in Parliament shall be transacted in Hindi or in English

[Jaska]

Ph2ter:

Some discipline can give wrong results, because it can be based on false premises and subsequent false conclusions. It can have false foundations. So, in such case the results/interpretations can disagree. And they can study the same object from different perspectives.

Can you give any example of such false premises and conclusions? 
How can they study the same object? Only if you define object so widely that it includes the objects of individual disciplines.

For example: you can say that the object is a group of ancient people, including their language, culture, and DNA. But still such a shared object is actually a sum of different objects, approachable only by different disciplines. There is no one discipline which could study all these parts of this total object.

[jdean]

And using my area (Wales) as another example of language change.

200 yrs ago there are documented examples of Welsh monoglots in Monmouthshire, SE Wales

Today nobody speaks Welsh as a first language here and even in areas of Wales (further West) where Welsh is still the first language everybody can also speak English.

[JonikW]

I agree with Jaska that in theory the results of the disciplines we're talking about here by their very nature can't disagree (they are all simply facts). The way we assess the results and draw conclusions from them very obviously can.

The way I see it, the problem is that our data in all these fields is limited and our opinions are always to an extent subjective. Hence the disagreements even between linguists or between geneticists or historians on the interpretation of so many of the limited facts we have.

I’ve come to agree that language can’t be predicted from DNA (kidnap a baby and move them to another culture, to give one example that I wouldn't recommend). At the same time I'd be amazed if we were to find skeletal remains of an I1 individual (to take a subject I know) anywhere in Europe in 10 CE who was also of exclusively NW European autosomal ancestry but who didn't turn out to be a speaker of a Germanic language or whose very recent ancestors weren't speakers of one if we could magically lift the lid on all the facts.

To that extent, at least, I think language can be seen in the genes. Probably the best approach for the fields is indeed to pool resources and try to work from there. 

There will always be egos though and I doubt whether these competing fields are capable of reaching a useful consensus most of the time because we all like to feel we're right. The other person has devoted their time to the less informative discipline but we picked exactly the right one.

[Cejo]

Cejo:

To me, subjecting hypotheses to evidence from a broad range of disciplines is often critical in achieving the rigor demanded by the scientific method. Restricting inquiry based on discipline/method is in some regards antithetical to what scientific means to me. In my view, an interdisciplinary viewpoint helps when it comes to the inductive reasoning (hypothesis-forming) phase of the scientific method, and when trying to avoid cognitive biases when analyzing results.

Interdisciplinary is of course important, too. But it requires by definition acknowledging the limitations of disciplines – it does not mean that language could be seen from the DNA or vice versa. Unfortunately many people who claim to be interdisciplinary are actually only unscientific, when they claim to achieve scientific knowledge by unscientific methods. The risk of cognitive bias is also present in interdisciplinary work.

Cejo:

Basically, I evidently don't attach the same meaning to "scientific" as you do, though I am a physical scientist, which may skew my viewpoint.

To me it seems that your definition agrees with mine. What do you think now?

Yeah, this is a good clarification. I think when you said that a discipline can only study its own subject, I interpreted it differently than you may have meant it, based on the above.